A new study (via Wired) from DuckDuckGo, a rival of Google, charges that Google search personalization is contributing to “filter bubbles.” Disputes this, Google says that search personalization is mostly a myth.
Filter bubbles in social media or search has been something of a controversy for a number of years now by Eli Pariser, who described it as how relevance algorithms tend to reinforce users’ existing beliefs and biases.
According to the study, 87 U.S. adults did searches for “gun control,” “immigration,” and “vaccinations” at the same time on June 24, 2018. These searches were conducted in both incognito mode as well as non-private-private mode. A majority of the searches were performed on the desktop, and a smaller percentage was done on mobile devices. This study was small in terms in the number of participants and query volume.
Below are the top-level findings according to DuckDuckGo’s discussion:
- Most people saw results unique to them, even when logged out and in private browsing mode.
- Google included links for some participants that it did not include for others.
- They saw significant variation within the News and Videos infoboxes.
- Private browsing mode and being logged out of Google offered almost zero filter bubble protection.
The DuckDuckGo post offers more in-depth discussion of the findings, as well as the raw data for download.
Comparing the variation in search results